Macrocyclic Lactam Synthesis via a Ring Expansion Reaction

The cripowellin ring skeleton, a macrocyclic [2.3.5]-bicyclic ketolactam, was .... Sean Henry , Ria Kidner , Mary R. Reisenauer , Igor V. Magedov , Ro...
0 downloads 0 Views 93KB Size
ORGANIC LETTERS

Macrocyclic Lactam Synthesis via a Ring Expansion Reaction: Construction of the Cripowellin Skeleton

2005 Vol. 7, No. 6 1031-1034

Bongjin Moon,*,†,‡ Sangbae Han,† Yeojin Yoon,† and Hyejin Kwon† Department of Chemistry, Sogang UniVersity, Seoul 121-742, Korea, and Center for BioactiVe Molecular Hybrids, Yonsei UniVersity, Seoul 120-749, Korea [email protected] Received December 3, 2004

ABSTRACT

The cripowellin ring skeleton, a macrocyclic [2.3.5]-bicyclic ketolactam, was smoothly generated via construction of a spiro(benzazepincyclohexane-1,3-dione) employing oxidative cyclization as a key step and a subsequent ring expansion reaction.

Cripowellins A (1) and B (2) (Figure 1) are members of the Amaryllidaceae alkaloids and were isolated from the bulbs of Crinum powellii by Velten and colleagues in 1997.1,2 Their structures were determined unambiguously by extensive spectroscopic methods including X-ray crystallographic analysis of cripowellin A diacetate (C14 and C6′ acetates).

Figure 1. Structures of cripowellins A (1) and B (2).

The absolute stereochemistry was postulated on the assumption that the carbohydrate moiety is biogenetically derived from β-D-glucose.1 In addition to their interesting insecticidal † ‡

Sogang University. Yonsei University.

10.1021/ol047520+ CCC: $30.25 Published on Web 02/16/2005

© 2005 American Chemical Society

activity, the 10-membered fused lactam skeleton of the cripowellins appears to be unique to the Amaryllidaceae alkaloid series. Careful structural examination shows that the main skeleton of cripowellin is closely related to other Amaryllidaceae alkaloids such as mesembrine (3), crinane (4, unnatural product), and crinine (5) (Figure 2). Oxidative cleavage of the C2-C3 bond of the tetrahydroisoquinoline moiety in the crinane skeleton reveals a structure that is equivalent to the cripowellin skeleton. From this point of view, it would also be interesting to determine if cripowellin is biosynthetically derived from other Amaryllidaceae alkaloids such as those of the crinane family. Our interest in a total synthesis of cripowellin led us to envision that the skeleton could be constructed via a ring expansion reaction of spiro(benzazepine-cyclohexane-1,3dione) intermediate 6, as shown in Figure 3. Intramolecular attack of the benzazepinylamine to the more electrondeficient ketone (the R-hydroxy ketone) in 6 should promote (1) Velten, R.; Erdelen, C.; Gehling, M.; Go¨hrt, A.; Gondol, D.; Lenz, J.; Lockhoff, O.; Wachendorff, U.; Wendisch, D. Tetrahedron Lett. 1998, 39, 1737. (2) Gehling, M.; Goehrt, A.; Gondol, D.; Lenz, J.; Lockhoff, O.; Moeschler, H.; Velten, R.; Wendisch, D.; Andersch, W.; Erdelen, C.; Harder, A.; Mencke, N.; Turberg, A.; Wachendorff-Neumann, U. German Patent 19610279, 1997.

Figure 4. Ban’s macrolactam syntheses via ring expansion.3

Figure 2. Comparison of the cripowellin skeleton with other Amaryllidaceae alkaloids.

a concomitant ring expansion to the keto-macrolactam ring 7. We hoped that the formation of the thermodynamically more stable amide functionality would provide a driving force for the formation of the fairly strained macrolactam skeleton of 7.

Because, to our knowledge, no similar reaction employing a cyclic amino group and a cyclohexane-1,3-dione system has been explored so far, we thought it would be interesting to see whether the key reaction shown in Figure 3 could be achieved. To prepare key substrate 8, we planned to use an intramolecular oxidative coupling reaction between an aryl moiety and a cyclohexane-1,3-dione, a reaction that has been extensively studied by the Kita group.5 Kita reported that five- and six-membered spirobenzannulated compounds could be obtained in reasonable yields using phenyl iodide bis(trifluoroacetate) (PIFA) as an oxidant. Although sevenmembered ring formation was not reported in their study, we hoped that such a cyclization would occur under similar conditions. Synthesis of the desired substrate 10 is illustrated in Scheme 1. Alkylation of N-piperonyl-N-tosyl amide (15) with excess 1,2-dibromoethane in the presence of K2CO3 in DMSO yielded N-bromoethyl-N-piperonyl-N-tosylamide (16) in 91% yield. Because it is known that direct C2-alkylation of cyclohexane-1,3-dione with an alkyl halide is sluggish due to strong competition with O-alkylation, we chose Piers’ protocol,6 which employs alkylation of 1,5-dimethoxy-1,4cyclohexadiene (17) and acid-catalyzed hydrolysis of the resulting product 18 to afford the desired substrate 10.

Scheme 1.

Synthesis of 10

Figure 3. Synthetic plan for the cripowellin skeleton.

In this paper, we describe synthetic efforts toward spiro(benzazepine-cyclohexane-1,3-dione) intermediate 8 from N-piperonyl-N-[2-(2,6-dioxocyclohexyl)ethyl]-toluenesulfonyl amide (10) or N-piperonyl-N-[2-(6-hydroxy-2-methoxyphenyl)-ethyl]toluenesulfonyl amide (11) and a model study of the ring expansion reaction with an N-toluenesulfonylprotected version of dideoxy derivative 8 (compound 19). A literature search led to our discovery of various ring expansion reactions of 2-aminoethyl-substituted cyclopentane-1,3-diones that were studied by Ban and colleagues.3 As shown in Figure 4, medium-sized ketolactams 12-14 could be obtained through controlled crisscross annulation, involving an acyclic amino alkyl group in the substrates.4 (3) Ohnuma, T.; Nagasaki, M.; Tabe, M.; Ban, Y. Tetrahedron Lett. 1983, 24, 4253. 1032

With compound 10 in hand, oxidative cyclization using various oxidants was attempted (Scheme 2). Unfortunately, (4) Ohnuma, T.; Tabe, M.; Shiiya, K.; Ban, Y. Tetrahedron Lett. 1983, 24, 4249. (5) Arisawa, M.; Ramesh, N. G.; Nakajima, M.; Tohma, H.; Kita, Y. J. Org. Chem. 2001, 66, 59. (6) Piers, E.; Grierson, J. R. J. Org. Chem. 1977, 42, 3755.

Org. Lett., Vol. 7, No. 6, 2005

Scheme 2.

Various Cyclization Attempts of 10

no desired cyclized product 19 was observed under the typical reaction conditions (PIFA in CF3CH2OH) developed by Kita and colleagues,5 and only unidentifiable, complex mixtures were obtained. Trials at highly dilute conditions (∼0.001 M) did not result in any improvement. Changing the N-protecting group from p-toluenesulfonyl to trifluoroacetate also did not provide any cyclization product under the same conditions. Changing the oxidant to PIDA (phenyl iodide diacetate) in methylene chloride showed no improvement. When the substrate was subjected to a one-electron oxidant, Mn(OAc)3, the reaction also yielded a complex mixture where piperonal was identified as the major component. In this case, it seems that the generated cyclohexane1,3-dione radical abstracted the benzylic hydrogen via 1,5hydrogen abstraction followed by further oxidation of the resulting benzylic radical by Mn(OAc)3 to the corresponding tosyl iminium ion, which was transformed to piperonal upon hydrolytic workup. A similar 1,5-hydrogen abstraction under Mn(OAc)3 conditions has been described in the literature.7 Because the oxidative cyclization approach using the cyclohexane-1,3-dione substrate was problematic, we shifted our strategy toward using 2-alkyl-3-methoxy phenol 11. Kita and colleagues have reported an intramolecular phenolic coupling reaction using a hypervalent iodine(III) reagent and showed that the reaction is a powerful tool in the syntheses of many Amaryllidaceae alkaloids.8 Compound 11 was prepared in a straightforward manner as shown in Scheme 3. Methoxymethyl (MOM)-directed lithiation of 20 followed by alkylation with allyl bromide and subsequent ozonolysis provided aldehyde 21 in 63% yield over two steps.9 Reductive amination of aldehyde 21 with piperonylamine provided the corresponding secondary amine 22 in 73% yield. After protecting the amine with a p-toluenesulfonyl group (TsCl, TEA, 67%), the MOM group of the resulting tosyl amide 23 was removed by treatment with a catalytic amount of p-toluenesulfonic acid in refluxing 2-propanol10 to give the desired phenol substrate 11. (7) Citterio, A.; Fancelli, D.; Finzi, C.; Pesce, L. J. Org. Chem. 1989, 54, 2713. (8) (a) Kita, Y.; Takada, T.; Gyoten, M.; Tohma, H.; Zenk, M. H.; Eichhorn, J. J. Org. Chem. 1996, 61, 5857. (b) Kita, Y.; Arisawa, M.; Gyoten, M.; Nakajima, M.; Hamada, R.; Tohma, H.; Takada, T. J. Org. Chem. 1998, 63, 6625. (9) (a) Simas, A. B. C.; Coelho, A.; Costa, P. R. R. Synthesis 1999, 1017. (b) Pocci, M.; Bertini, V.; Lucchesini, F.; De Munno, A.; Picci, N.; Iemma, F.; Alfei, S. Tetrahedron Lett. 2001, 42, 1351.

Org. Lett., Vol. 7, No. 6, 2005

Scheme 3.

Synthesis of 11

In contrast to the use of cyclohexane-1,3-dione 10, when phenol substrate 11 was treated with PIFA in CF3CH2OH at room temperature, the oxidatively cyclized spirobenzazepin product 24 was obtained in a moderate yield (58%) (Scheme 4). Having key intermediate 24 in our hands, we next

Scheme 4.

Cyclization of 24

investigated the functionalization of the 5-methoxycyclohexadienone moiety to install the C-14 and C-15 oxidation state of cripowellin at this early stage. Unfortunately, various conditions, including hydrolysis (HCl/H2O), epoxidation (tBuOOH/Triton B11 or dimethyldioxirane12), and reduction (H2, PtO2), failed to give any identifiable products. In most cases, the dienone decomposed under these conditions. After (10) Miyashita, N.; Yoshikoshi, A.; Grieco, P. A. J. Org. Chem. 1977, 42, 3772. (11) (a) Yang, N. C.; Finnegan, R. A. J. Am. Chem. Soc. 1958, 80, 5845. (b) Payne, G. C. J. Org. Chem. 1960, 25, 275. (12) (a) Adam, W.; Bialas, J.; Hadjiarapoglou, L. Chem. Ber. 1991, 124, 2377. (b) Murray, R. W.; Singh, M. Org. Synth. 1997, 74, 91. 1033

extensive screening of various reagents, only Stryker’s reagent [(Ph3PCuH)6]13 was found to successfully mediate a clean reaction to provide the reduced product 25 in good yield (81%). Fortunately, 1,4-conjugate reduction was favored over 1,6-conjugate reduction in this case. In contrast to cyclohexadienone 24, hydrolysis of methyl enol ether 25 proceeded cleanly to give cyclohexanedione 19 in 72% yield. Finally, we treated spiro(benzazepin-cyclohexanedione) N-tosyl amide 19 with sodium naphthalenide solution14 to remove the toluenesulfonyl group, expecting either the amine form of 26 or hemiaminal form of 27 as the product. The starting material disappeared in 10 min and formed a single new product, as judged by TLC analysis. In contrast to the plane-symmetric dione 19, the resulting product lost significant symmetry, according to 1H NMR spectroscopic analysis. Surprisingly, additional spectroscopic analyses, including IR, 13 C NMR, COSY, HMBC, and HR-MS (ESI), of the product revealed that the product was not the predicted intermediate hemiaminal 27 but the desired ketoamide 9. The IR spectrum of the product showed the characteristic bands for the C17 (13) (a) Mahoney, W. S.; Brestensky, D. M.; Stryker, J. M. J. Am. Chem. Soc. 1988, 110, 291. (b) Chiu, P.; Li, Z.; Fung, K. C. M. Tetrahedron Lett. 2003, 44, 455. (14) (a) McIntosh, J. M.; Matassa, L. C. J. Org. Chem. 1988, 53, 4452. (b) Heathcock, C. H.; Blumenkopf, T. A.; Smith, K. M. J. Org. Chem. 1989, 54, 1548.

1034

ketone stretch at 1696 cm-1 and the C13 amide CdO stretch at 1635 cm-1, which closely correlate with the values observed for cripowellin A (1692 and 1653 cm-1, respectively1). The presence of the ketone and amide functionality was also confirmed by 13C NMR and HMBC (δ 213.8 and 173.5, respectively). In summary, we have demonstrated that the cripowellin skeleton can be efficiently constructed by ring expansion of spiro(benzazepine-cyclohexane-1,3-dione) 19, which was obtained by intramolecular oxidative cyclization of phenolic compound 11 followed by a reduction/hydrolysis sequence. We are currently pursuing functionalization of compound 9 to achieve the total synthesis of cripowellin and development of an asymmetric version of this synthetic strategy. Acknowledgment. This work was supported by the Korea Research Foundation (KRF-2002-003-C00079). The authors thank Professor Hanbin Oh for ESI-FTMS analysis of compounds 25 and 9 and Dr. Michael Haukaas for his editorial contributions to the manuscript. Supporting Information Available: Experimental procedures and characterization data for all new compounds. This material is available free of charge via the Internet at http://pubs.acs.org. OL047520+

Org. Lett., Vol. 7, No. 6, 2005